A Review of Normal Values of Infant Sleep Polysomnography

Transcription

1 Pediatrics and Neonatology (2013) 54, 82e87 Available online at journal homepage: REVIEW ARTICLE A Review of Normal Values of Infant Sleep Polysomnography Daniel K. Ng*, Chung-Hong Chan Department of Paediatrics, Kwong Wah Hospital, Hong Kong SAR, China Received Aug 28, 2012; received in revised form Oct 28, 2012; accepted Dec 28, 2012 Key Words apnea; infant; neonate; polysomnography Background: The objective of this study was to summarize current information about the normal values on infant sleep polysomnography for clinical use. Methods: MEDLINE (Ovid), EMBASE (Ovid), and CINAHL (Ovid) from January 1976 to May 2007 were searched. Two reviewers independently reviewed all relevant articles, using preset inclusion criteria. The population of interest included children aged less than 1 year. Studies in infants with known major anomalies were excluded. The results on apneas were extracted and analyzed. Results: For obstructive apnea, the upper limit of normal values was less than 1.0 per hour, and for mixed apnea, the current data suggested the upper limit of normal values was less than 1.0 per hour. For central apnea defined as cessation of respiratory efforts for more than 3 seconds, the current data suggested that the upper limit of the normal central apnea index was 45 per hour for 1-month-old infants, 30 per hour for 2-month-old infants, 22 per hour for 3-month-old infants, and between 10 and 20 for the older age groups. For the desaturation episode defined as SpO 2 less than 80% for any length of time, the current data suggested the upper limit of normal values to be 14.7 episodes per hour for day 1, 41 episodes for day 4, and 15.1 episodes for day 39. Conclusion: The normal values of obstructive apnea, mixed apnea, and central apnea are well established for neonates and infants. With these normal values, sleep polysomnography study should be routinely used to quantify the severity of breathing disorders during sleep in those neonates at risk for these disorders. Copyright ª 2012, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved. * Corresponding author. Department of Paediatrics, Kwong Wah Hospital, Waterloo Road, Hong Kong SAR, China. address: dkkng@ha.org.hk (D.K. Ng) /$36 Copyright ª 2012, Taiwan Pediatric Association. Published by Elsevier Taiwan LLC. All rights reserved.

2 Normal value of infant PSG Introduction After delivery, continuous breathing pattern becomes essential, and it represents a dramatic departure from fetal life when the breathing movement is irregular. Hence, it is not surprising that apnea is a common phenomenon during early infancy. During infancy, apnea is commonly defined as breathing cessation that lasts for more than 20 seconds or as a briefer episode if associated with bradycardia or pallor or cyanosis. 1 In the field of neonatology, apnea is often not routinely differentiated into central or mixed or obstructive apnea. The use of polysomnography (PSG) with simultaneous assessment of multiple relevant variables would help differentiate the types of apnea. This differentiation might well be important for shedding clues as to the underlying etiologies and prescription of proper treatments, e.g., caffeine for significant central apnea and continuous positive airway pressure for significant obstructive apnea. Obstructive sleep apnea was found in 58% of infants referred to a single center for apparently life-threatening events (ALTEs), 2 and a population-based case-control study found that nieces/nephews of patients with obstructive sleep apnea syndrome (OSAS) were twice as likely as controls to have offspring who died of sudden infant death syndrome (SIDS). 3 However, an extensive review was published to address questions about SIDS and its possible relationship with OSAS and found no strong evidence. 3 Hence, the relationship between ALTEs or SIDS and OSAS remains controversial. To cast light on this controversy, sleep PSG appears to be a useful tool. However, the use of PSG in detecting sleep-related apneas in infants remains difficult because of the different definitions of apneas used in different studies, which has resulted in the current lack of widely accepted normal values for different parameters recorded in PSG of infants. In the current review, we analyzed the available data in the literature on infant sleep PSG parameters in healthy infants. The primary objectives were (1) to provide information about the available normal data on infant sleep PSG for clinical use and (2) to review the technical specifications for infant PSG. 2. Methods 2.1. Sources The following bibliographic databases were searched: MEDLINE (Ovid), EMBASE (Ovid), and CINAHL (Ovid), limited to studies conducted from January 1976 to May The keywords used for search included neonate, infant, PSG, normal, control, healthy, and apnea. The search strategy is available on request. The search was conducted in May Standard textbooks in this area were also handsearched for additional data. Bibliographies of review articles, systematic reviews, and retrieved articles were also searched for candidate articles. Non-English candidate articles were translated by a web-based translation engine (Systran, systransoft.com/index.html) Study selection Case-control, cohort, and epidemiological studies were retrieved for analysis. Case reports and letters to editors were not retrieved for analysis. For multiple publications of the same material, only one article was chosen. Two reviewers (CCH and DKN) independently reviewed all pertinent articles, using preset inclusion criteria. Discrepancies were resolved by consensus. Reviewers were not blinded to the source of publication or authors at any step during study selection Inclusion criteria The population of interest included children aged less than 1 year. Studies in infants with known major anomalies (e.g., Down syndrome) were excluded. In this review, normal infants included those who were born at term after a normal gestation, were not suffering from SIDS or ALTEs, and have no family history of sleep apnea, SIDS, or ALTEs. PSG or pulse oximetry was performed in all infants. The methods of measurement and the number of polysomnographic parameters in all included studies were listed. Other confounding factorsdi.e., gender of infants, age of infants, sources of recruitment, and nationalitiesdwere also listed for each included study. 3. Results A total of 147 articles met our initial search criteria. Ten met our primary inclusion criteria. The characteristics of the included studies are listed in Table The configuration of infant PSG Various signals were recorded to monitor a wide range of electrophysiological functions in infants during sleep. Most studies used similar configurations to perform infant PSG (Table 2). Electroencephalogram (EEG), electrocardiogram (ECG), electro-oculogram (EOG), thoracic and abdominal efforts, and oronasal thermistor were used by Kahn, 4 Schluter et al, 5 Franco et al, 6 Rebuffat et al, 7 Kato et al, 8 and Guilleminault et al. 9 Submental and leg electromyogram (EMG) was used by Schluter et al, 5 Franco et al, 6 Guilleminault et al. 9 Either transcutaneous oxygen monitor or pulse oximeter was used to measure oxygen saturation. Kahn 4 and Schluter et al 5 recorded both transcutaneous oxygen tension and pulse oximetry. Actigraph was recorded by Kato et al, 8 Kahn, 4 Rebuffat et al, 7 and Franco et al 6 to detect infant gross body movement. In the study by Schluter et al, 5 video recording was added Definition of apneas Contrary to the commonly used definition of apnea in neonatology, 1 apnea was generally defined to last 3 seconds (Table 3). Obstructive apnea was defined as a continuous activity from a strain gauge that detects respiratory efforts and flat tracing, i.e., no air movement. Central apnea was defined as flat tracings

3 84 D.K. Ng, C.-H. Chan Table 1 Characteristics of studies using polysomnography (PSG) to assess normal values of apneas in normal healthy infants. First author, year Country Sample PSG and sleep laboratory setting Schluter, 2001 Germany 354 male and 327 females 1e24 months old Median gestational age: 39 weeks old (38e40 weeks old) Kato, 2000 Belgium 470 males and 553 females 2e27 weeks old Guilleminault, 1981 United States Age and number of infants 3 wk: 10 6 wk: 10 3 mo: mo: 10 6 mo: 6 Franco, 2000 Belgium 11 boys and 9 girls Median age: 11.5 wk Rebuffat, 1994 Belgium 8 normal infants, age/gender: not stated Kahn, 2000 Belgium 1135 normal infants classified into different age groups 1 night (time and duration not stated) Ambient temperature: 20 2 C Feeding: one feeding and postprandial period provided 1 night (8 hours, 2100e0500 hours) Ambient temperature: 20e23 C 1 night (24 hours, 1100e1100 hours) Ambient temperature: not stated Feeding: not stated 1 night (8 hours, timing not stated) Ambient temperature: 20 C, 25 C, 30 C Sleep position: supine 1 night (2100 to 0700 hours) Ambient temperature: 20e23 C Humidity: 30% Sleep position: supine 1 night (time not stated, at least 360 min ) Ambient temperature: not stated obtained simultaneously from strain gauges and thermistors. Mixed apnea was defined as central apneas followed. The study by Guilleminault et al 9 separated apneas by duration. The study by Kahn 4 defined central apnea as apnea longer than 4 seconds. Kahn 4 also stated that obstructive events due to displacement of the thermistors, any doubtful episodes, such as obstructed breathing preceded by a movement or a sigh, were rejected Age-related changes in the normal values of apnea index The studies conducted by Kato et al, 8 Schluter et al, 5 and Kahn 4 evaluated the age-related changes in normal values of polysomnographic variables in infants. All studies reported the normative data on obstructive apnea index (OAI) and mixed apnea index (MAI). However, the study by Kato et al 8 did not provide the complete data for central apnea. Downward trends with increasing age were noted in the 90 th percentile and 75 th percentile curve of obstructive apnea and mixed apnea. Both obstructive apnea and mixed apnea indices were highest in the neonatal period. The duration of obstructive and mixed apneas also tended to decrease with age. The frequency of central apnea decreases with age, but the duration was steady at around 3e5 seconds across all age groups Normal values of obstructive apnea In the study by Schluter et al, 5 themedianvalueofoai was zero in all age groups. In the study by Kato et al, 8 however, the median OAI was 0.15 per hour for 2-weekold infants and dropped to zero by the time they are 8 weeks old. For the upper limit of OAI, both studies by Schluter et al 5 and Kato et al 8 showed a similar decreasing trend as the infants got older. The 90 th percentile of OAI was to per hour from 2 weeks onward, and decreased to a level of to perhourby2e3months of age. Another study by Kahn 4 presented the age-specific rates of obstructive apnea and mixed apnea per hour of total sleep up to 27 weeks. The median values of OAI were zero across all age groups. However, the upper limit (90 th percentile) was not provided in Kahn s study. It confirmed the previous observation that pure obstructive apnea was rarely seen in infants.

4 Normal value of infant PSG 85 Table 2 First author, year Schluter, 2001* Kato, 2000 Guilleminault, 1981 Franco, 2000 Rebuffat, 1994 Kahn, 2000 Configuration of PSG. Different configurations were used during the study period, with following channels: 6 EEG channels, EOG, EMG, thoracic and abdominal breathing movements by plethysmography, oronasal airflow (method of measurement was not mentioned), transcutaneous blood gases, oxygen desaturation as measured by pulse oximeter and ECG. Two EEG with central and occipital leads (C4/A1 and C3/A2), two EOG and ECG. Thoracic respiratory movements were measured by impedance and airflow with thermistors taped under both nostrils and on the side of the mouth. Oxygen saturation was recorded continuously by a transcutaneous sensor (Nellcor, USA). Gross body movements were measured using an actigram placed on one arm. The data were collected on a computerized infant sleep recorder (Alice Recording System III, Healthdyne, USA). EEG (C3/A2-C4/A1), EOG, Digastric EMG, ECG lead, Holter ECG, abdominal and thoracic strain gauges, nasal and oral thermistors, nasal catheter for measurement of percent of CO 2 in expired gases, ear oximeter, skin surface oxygen electrode to measure oxygen tension. 2 EEG channels, 2 EOG, digastric EMG, Thoracic respiratory movements were measured by impedance and airflow with thermistors taped under both nostrils and on the side of the mouth. Gross body movements were measured using an actigram placed on one arm. Oxygen saturation was recorded continuously by a transcutaneous sensor. The data were collected on a computerized infant sleep recorder (Morpheus System, Medatec, Brussels, Belgium). EEG, EOG, ECG, Thoracic respiratory movements were measured by impedance and airflow with thermistors taped under both nostrils and on the side of the mouth. Gross body movements were measured using an actigram placed on one arm. Oxygen saturation was recorded continuously by a transcutaneous sensor. Motion artifacts were detected from body movements and the analysis of the oximeter waveform. The data were collected on standard polygraph records (paper speed 10 mm/s) or on computerized systems (Alice III, Apreco, Belgium). EEG, EOG, ECG, thoracic respiratory movements were measured by impedance and airflow with thermistors taped under both nostrils and on the side of the mouth. Gross body movements were measured using an actigram placed on one arm. Oxygen saturation was recorded continuously by a transcutaneous sensor. Motion artifacts were detected from body movements and the analysis of the oximeter waveform. The data were collected on standard polygraph records (paper speed 10 mm/s) or on computerized systems. Table 3 Definition of apnea in studies. First author, year Apnea OA CA MA Schluter, 2001* 3 s Cessation of oronasal airflow Cessation of Not clear with ongoing respiratory efforts respiratory efforts Kato, s Continuous deflections from Flat tracings from both the strain gauges and the thermistors Guilleminault, 1981 Classified into 3e6 s 6e10 s 10e15 s >15 s Continuous deflections from Franco, s Continuous deflections from Rebuffat, s Continuous deflections from Kahn, s for OA and MA >4 for CA Continuous deflections from CA Z central apnea; MA Z mixed apnea; OA Z obstructive apnea. * Different setups of PSG were used during the study period. Flat tracings from both the strain gauges and the thermistors Flat tracings were obtained simultaneously from strain gauges and thermistors Flat tracings were obtained simultaneously from strain gauges and thermistors Flat tracings were obtained simultaneously from strain gauges and thermistors Central apnea followed

5 86 D.K. Ng, C.-H. Chan 3.5. Normal values of mixed apnea For mixed apnea, the median value of MAI approached zero across all age ranges in infants studied by Kato et al 8 and Schluter et al. 5 For the 90 th percentile, both studies showed similar results, i.e., the MAI was 0.3e0.5 per hour at 2 weeks of life and decreased to 0.2e0.4 per hour by 2e3 months of age. The study by Kahn 4 showed that median values of MAI were zero across all age groups, similar to OAI. However, the upper limit (90 th percentile) was not provided in Kahn s study Normal values of central apnea Schluter et al 5 reported the median central apnea index (CAI) of infants across the neonates from 1 month to 12 months to be 5e10 per hour; they further noted that the 95 th percentile of CAI was 45 for 1-month-old neonates, 30 for 2-month-old neonates, 22 for 3-month-old neonates, and between 10 and 20 for older age groups Effect of ambient temperature of sleep laboratory on frequency of apneas Only one study (conducted by Franco et al 6 ) evaluated the effect of ambient temperature on frequency of apneas in normal infants. Each infant studied was exposed to the ambient temperatures of 20 C, 25 C, and 30 C within 1 night. They found that the CAI during rapid eye movement (REM) sleep was significantly elevated in higher temperatures but not in non-rem sleep. All other frequencies of respiratory events were similar in either REM or non-rem sleep First night effect Only one study 7 used PSG to evaluate the night-to-night variability in apnea indices in normal infants. Out of eight normal infants studied by Rebuffat et al, 7 four had either mixed or obstructive apneas during the 1 st night. The apnea indices remain comparable in the 2 nd night and 3 rd night Circadian pattern of distribution in mixed and obstructive apneas in normal infants The study by Guilleminault et al 9 reported the circadian pattern of mixed and obstructive apneas in normal infants. In normal infants, abnormal respiratory events during sleep were more likely to occur between 1 AM and 6 AM, and more than 40 minutes after sleep onset Oximetry Four studies by Stebbens et al, 10 O Brien et al, 11 and Poets et al 12,13 reported the median baseline pulse oximetry (SpO 2 ) and desaturation (defined as SpO 2 less than 80%) of infants using pulse oximetry in the beat-to-beat mode. The gestational age of infants was at least 37 weeks, and their median age ranged from 1 hour 11 to 191 days 13 of life. The median SpO 2 ranged from 97.6% (92e100%) 12 to 98.3% (88.7e100%) 11 during the 1 st month of life. From 5 weeks to 6 months, the SpO 2 ranged from 99.6% (97.9e100%) to 99.9% (98.6e100%). 13 The median rate of desaturation was 0 per hour (range: 0e14.7) at 1 hour of life. 11 On day 4 of life, the median rate of desaturation was also zero (range: 0e41 per hour). 12 The median rate of desaturation was 0.9 per hour (range: 0e15.1) for infants on day 39 of life. 10 Only one study 11 reported the median duration of desaturation with and without apnea separately. The median duration of desaturation without concomitant apnea at 1 hour of life was 9.3 seconds (0.5e286.8 seconds), and that with apnea was 0.8 seconds (0.3e89.6 seconds). 11 The median duration of desaturation decreased from 5.1 seconds (4.2e9.9 seconds) at 4 days of life 11 to 0.9 seconds (0.4e1.8 seconds) at 3 months of age. 12 Forty-eight percent of desaturation was associated with concomitant apnea at 1 hour of life. 11 Desaturation episodes were more often associated with apnea with increasing age, and the proportion of desaturation episodes associated with apnea increased from 48% at 1 hour of life 11 to 93.8% at 6 months of age Discussion In neonatology, apnea is often defined as cessation of breathing lasting more than 20 seconds. 1 However, the normal value for apnea so defined is not available. Furthermore, recent evidence suggested that obstruction of the airway was frequently present in the so-called central apneas defined as longer than 20 seconds. 14 Idiong et al 15 analyzed >4000 apneas in preterm neonates and found that central apneas occurred predominantly with durations of less than 10 seconds, whereas apneas >10 s were usually mixed apneas. It has not been a common practice to further classify apneas into central, obstructive, and mixed apneas. For the proper management of apneas, it would be important to make the distinction even though the majority (70e80%) of apneas in the neonatal period would be central apnea. 16 PSG would be the ideal tool to make the distinction and to quantify the abnormalities. The current review summarized the available data about sleep PSG in healthy full-term neonates and infants. For obstructive apnea, it was clear from current data that the upper limit of normal values was less than 1.0 per hour, and neonates with obstructive apnea 1.0 would be highly suggestive of having a disease that warrants investigations for the site(s) of obstruction. Data are currently lacking in the natural course of these obstructive apneas, and further research is urgently needed. For mixed apnea, the current data suggested that the upper limit of normal is less than 1.0 per hour, and those with mixed apnea 1.0 per hour are suggestive of a disease state warranting further investigations along the line of obstructive apnea. For central apnea defined as cessation of respiratory efforts for more than 3 seconds, the current data suggested that the upper limit of normal CAI was 45 for 1-month-old infants, 30 for 2-month-old infants, 22 for 3-month-old infants, and between 10 and 20 for older age groups. For the desaturation episode defined as SpO 2 less than 80%, the current data suggested the lower limit of normal values to be 14.7 episodes per hour for day 1, 41

6 Normal value of infant PSG 87 episodes for day 4, and 15.1 episodes for day 39. It is important to realize that the current data were obtained from pulse oximeter using a beat-to-beat mode, whereas most commercial pulse oximeters used averaging of data. Hence, the normal values that are reported may not be directly applicable to patients monitored with a different mode. Another important parameter about desaturation would be the percentage of time spent in each saturation zone (e.g., SpO 2 >90%, between 80% and 90%). Unfortunately, there are currently no normal data on this very important parameter. In conclusion, this review summarizes the available information on the normal values of sleep PSG data of normal full-term infants for use by clinicians. With these normal values, sleep PSG study should be routinely used to quantify the severity of breathing disorders during sleep in those neonates at risk for these disorders. The normal oximetry values warrant further research as the current data addresses only the number of episodes of SpO 2 less than 80% for any length of time. This addresses neither the length of episode nor the normal value of percentage of time spent in SpO 2 between 80% and 90%. References 1. Task Force on Prolonged Apnea. American Academy of Pediatrics. Prolonged apnea. Pediatrics 1978;61:651e2. 2. Committee on Fetus and Newborn. American Academy of Pediatrics. Apnea, sudden infant death syndrome, and home monitoring. Pediatrics 2003;111:914e7. 3. Guilleminault C, Pelayo R, Leger D, Philip P, Ohayon M. Sleepdisordered breathing and upper-airway anomalies in firstdegree relatives of ALTE children. Pediatr Res 2001;50:14e Kahn A. Breathing during sleep in infancy. In: Loughlin GM, Caroll JL, Marcus CL, editors. Sleep and breathing in children: a developmental approach. 1st ed. New York: Marcel Dekker; p. 405e Schluter B, Buschatz D, Trowitzsch E. Perzentilkurven polysomnographisher parameter für das erste und zweite Lebensjahr. Somnologie 2001;5:3e Franco P, Sqliwowski H, Dramaix M, Kahn A. Influence of ambient temperature on sleep characteristics and autonomic nervous control in healthy infants. Sleep 2000;23:401e7. 7. Rebuffat E, Groswasser J, Kelmanson I, Sottiaux M, Kahn A. Polygraphic evaluation of night-to-night variability in sleep characteristics and apneas in infants. Sleep 1994;17:329e Kato I, Franco P, Groswasser J, Kelmanson I, Togari H, Kahn A. Frequency of obstructive and mixed sleep apnea in 1023 infants. Sleep 2000;23:487e Guilleminault C, Ariagno R, Korobkin R, Coons S, Owen- Boeddiker M, Baldwin R. Sleep parameters and respiratory variables in near miss sudden infant death syndrome infants. Pediatrics 1981;68:354e Stebbens VA, Poets CF, Alexander JR, Arrowsmith WA, Southall DP. Oxygen saturation and breathing patterns in infancy: 1. Full term infants in the second month of life. Arch Dis Child 1991;66:569e O Brien LM, Stebbens VA, Poets CF, Heycock EG, Southall DP. Oxygen saturation during the first 24 hours of life. Arch Dis Child Fetal Neonatal Ed 2000;83:F35e Poets CF, Stebbens VA, Lang JA, O Brien LM, Boon AW, Southall DP. Arterial oxygen saturation in healthy term neonates. Eur J Pediatr 1996;155:219e Poets CF, Stebbens VA, Southall DP. Arterial oxygen saturation and breathing movements during the first year of life. J Dev Physiol 1991;15:341e Rigatto H. Breathing and sleep in preterm infants. In: Loughlin GM, Caroll JL, Marcus CL, editors. Sleep and breathing in children: a developmental approach. 1st ed. New York: Marcel Dekker; p. 495e Idiong N, Cates D, Lemke R, Alvaro RE, Kwaitkowski K, Rigatto H. Profile and significance of the various types of apneas in preterm infants (abstract). Pediatr Res 1996;39: 328A. 16. Ng DK, Leung LC, Tong TF, Chan CH, Wong SF. Airway obstruction in premature newborns: a missing link. Pediatrics 2005;115:1789.